AI/ML Engineer · Biomedical Technology
RAG applications, biomedical signal and imaging pipelines, model evaluation, APIs, and packaging for real use. Research and portfolio work only where clinical context applies—not medical devices.
Focus
RAG systems, document intelligence, and LLM-powered workflows.
Classification, prediction, anomaly detection, medical imaging pipelines, and rigorous evaluation.
Analytics pipelines, SQL analysis, dashboards, and decision support.
APIs, Docker, experiment tracking, CI/CD, and deployment.
Portfolio
RAG · NLP · Hugging Face · FAISS · Streamlit · Live Demo
Built and deployed a document intelligence app: upload PDFs/TXT, retrieve semantically relevant passages, and generate grounded answers with open-source language models.
End-to-end ML · Biomedical · CI/CD · Research / educational only
End-to-end arrhythmia classification on MIT-BIH: Random Forest and 1D-CNN under inter-patient evaluation, MLflow tracking, FastAPI inference, Docker, and CI. Multi-model benchmark path runs offline on synthetic data.
Clinical RAG · LangChain · FAISS · Streamlit
RAG for biomedical documents: PDF ingestion, vector retrieval, and source-tracked answers. Designed as a portfolio demonstration of clinical document intelligence patterns.
Medical Imaging · DICOM/NIfTI · 3D + MPR · Next.js · Three.js · Research only
Format-agnostic CT pipeline: ingest DICOM or NIfTI, normalize to a canonical volume, validate, segment, export meshes, and explore organs in a synced 3D + multi-planar (axial/coronal/sagittal) viewer. Portfolio / learning use only — not a medical device.
More
DSP noise reduction, Random Forest predictive maintenance (synthetic data), and SQL calibration tracking. Fully runnable offline.
Generate, clean, and analyze retail sales data; answer region/category/customer questions; export CSVs for Excel/Power BI.
Runnable Random Forest failure prediction on synthetic industrial process features. Offline demo with clear metrics output.
Skills
Python, SQL, NumPy, Pandas
PyTorch, scikit-learn, experiment tracking
LangChain, FAISS, Hugging Face, embeddings
DICOM, NIfTI, Three.js, Next.js
FastAPI, Streamlit, Docker, CI/CD
About
B.Sc. Biomedical Technology. My work spans RAG applications, predictive modeling, biomedical signal processing, medical imaging pipelines, data analytics, APIs, and deployment. I take projects from raw data and experimentation through evaluation, backend integration, and packaging for use.